OPENPROOF — RPO v0.1
Signed JSON • Human-readable PDF • Public hash
RPO Playground · Open version

Try the RPO Playground — from narrative to draft proof bundle

Paste a short narrative, add a few identifiers, and generate a draft RPO JSON bundle with heuristic scores and a public hash. This page demonstrates how the RPO standard can structure a story into something verifiable.

Experimental
Deterministic
Open-source friendly
This is an open playground — no AI model, no registry, no legal value. It is here to show how an RPO bundle can look and behave, before plugging advanced engines or CNRS modules.
Play with a narrative

1. Input

Start with a short narrative (1–3 paragraphs). You can anonymise names and dates. The goal is to see how the structure behaves, not to judge the case.

Tip: include at least one date or time marker, a location, and 2–3 concrete actions.

2. Draft RPO bundle

Fill the form on the left and click “Generate” to see the bundle. 



        

          

            Scientific pilot (CNRS × TruthX)
          

          

            For an advanced psycho-medical-forensic analysis (interpretive
            coherence, narrative structure, markers of coercive control and
            narrative inversion), you can request access to the
            CNRS × TruthX research pilot.
            

            
              Apply for the CNRS pilot
            
          

        

      

    


    
    

      

        
How this page works

        

          This playground applies the minimal RPO “civil code” to your narrative.
          It does not use any AI model: everything runs locally in your browser.
        

        
    
          
  • 
            You provide a short narrative and basic identifiers (issuer, subject,
            title).
          
    • 
          
  • 
            The engine computes deterministic heuristics: sentences, evidence
            markers, anchors.
          
    • 
          
  • 
            A minimal rpo.json bundle is generated with a
            public_hash.
          
    • 
          
  • 
            The JSON is ready to be validated, hashed again, or integrated into
            your own tools.
          
    • 
        

        

          The structure mirrors the official RPO v0.1 specification, with
          additional fields marked as heuristic_scores and
          heuristic_anchors for research use.
        

      


      

        
What this engine does not do

        

          This open engine is deliberately limited. It is a demonstrator, not a
          forensic authority.
        

        
    
          
  • It does not determine truth or falsity.
    • 
          
  • 
            It does not replace legal analysis or expert
            assessment.
          
    • 
          
  • 
            It does not connect to any remote registry or
            blockchain.
          
    • 
          
  • 
            It does not embed psycho-forensic AI models in the
            open version.
          
    • 
        

        

          The goal is to make the structure of evidence transparent and testable,
          before connecting more advanced scientific engines in restricted pilots.
        

      

    


    
    

      

        Here you see the open RPO Playground.
        

        If you want to test the
        scientific / CNRS pilot version with advanced
        psycho-forensic modules and real cases, you can apply to the pilot
        programme.
      

      
        Apply for the CNRS pilot